Given that AI is top of mind for so many of us right now, I’ve started polling our in-house friends about how they’re using AI in their practice (here is the last set of anecdotes – up to 48 overall!). As one might expect, the perspectives and views are varied:
- “Sometimes I think AI has become my favorite colleague. Then it cites a regulation that doesn’t exist and “oops, there it is,” I remember why humans are still employed.”
- “The number of vendors pitching AI platforms to help craft disclosures at the last conference I went to was beyond mind-boggling. Half of those service providers won’t exist next year is my prediction. Bet on it.”
- “Our disclosure committee meetings haven’t gotten shorter due to using AI for prep. Some would say they’ve gotten better. Some would say why have the meetings at all in reality.”
- “I honestly think AI is following the same path as email. First it was controversial. Then it became accepted. Eventually it’ll just become infrastructure.”
- “Every vendor promises AI will revolutionize legal. I’ve lived through enough ‘revolutions’ to know that reality is usually somewhere in the middle.”
- “From conversations across the profession lately, many seem to be in a deliberate, early-innings phase – experimenting in low-risk areas while working through guardrails around confidentiality, privilege and data handling.
Honestly, the common use cases I hear about are the unglamorous but genuinely useful ones: first-draft summaries, organizing and comparing documents, surfacing relevant precedent and taking a first pass at routine correspondence that a lawyer then reviews and owns.
I’d expect adoption to increase meaningfully in the near term, but I think the key question stays the same: ‘who’s driving the car?’ The tools can give lawyers a faster starting point, but the lawyer stays accountable for the output.”
- “The first draft comes from AI. The judgment comes from lawyers. That’s a partnership I’m comfortable with. At least right now in 2026.”
- “We’re a Fortune 100 company so our biggest challenge isn’t whether AI works. It’s making sure thousands of people use it consistently, securely and responsibly. Technology problems are usually easier than change-management problems. I sometimes ask “what Bing Crosby would do?”
- “I don’t use AI because I think it’s smarter than I am. I use it because it never gets tired of comparing fifty versions of the same disclosure.”
- “I am so tired of talking about AI already.”
- “I’m cautiously optimistic. Every month it gets noticeably better. Every month I also find new ways it can be confidently wrong.”
- “Legal don’t usually get rewarded for resisting productivity improvements. We get rewarded for delivering sound advice quickly. AI happens to help with one of those things. The other one still belongs to us.”
Authored by
Broc Romanek